Ribosomes, the protein synthesising machines, keep translating until they reach a stop codon. Stop codons, unlike the previous mRNA codons before them do not code for amino acids ceasing protein production. Sometimes ribosomes misread stop codons and instead keep translating until the next one. What if a cell can exploit this read-through to aid in translation regulation? Yordanova et. al latest Nature paper1 may have uncovered just that.
A peak in the profile
Ribosomes are rather large complexes. This means that at any stage during translation they ‘hide’ around 30nt of mRNA. Whilst hidden these nucleotides are protected from the action of RNases (enzymes that degrade RNA). Ribosome profiling is a technique that exploits this protection to isolate the short nucleotide sequences that a ribosome is currently translating. Once sequenced the fragments are mapped onto the original gene to identify the location of the ribosome. Since ribosomes trundle along the length of the mRNA you would expect ribosomes to be found in all regions up to the stop codon (when the ribosome falls off) (Figure 1). However, some regions of mRNA show a higher fraction of ribosomes than others giving rise to peaks in the profile. These regions are proposed to be sites of ribosome stalling.
Yordanova’s team used results from ribosome profiling to analyse the gene AMD1. AMD1 encodes adenosylmethionine decarboxylase 1, an important enzyme in polyamine synthesis and neural differentiation. Unexpectedly, they observed a strong peak 384 nts downstream of the stop codon. This peak corresponded to a downstream stop codon, suggesting that a proportion of the ribosomes were translating through the standard stop codon and then stalling at the later one. This additional translation extension is referred to as the AMD1 tail.
Frogs have tails too
It is plausible that the ribosome profiling data could represent mRNA protection not related to translation; for example by RNA binding proteins in a nucleoprotein complex with a similar sedimentation value to ribosomes. However, the authors found the same spikes in the ribosome profiles of AMD1 from mice, rats, fish and frogs. Evolutionary convergence of the tailORF reflects its importance.
The tale of the tail
When the standard stop codon was replaced with a sense codon (to allow constitutive translation of the tail), it was found that production of AMD1 was almost completely lost. This was key evidence that the tail acts to repress translation. To quantify this more accurately the AMD1 tail sequence was cloned downstream and in frame with Renilla luciferase (a useful reporter of gene expression levels). A 65-fold drop in expression levels was seen when the tail was translated constitutively.
Further experiments indicated that the drop-in expression levels weren’t due to decreased stability of the extended protein. Instead, the team propose that ribosomes that infrequently readthrough the stop codon stall at the tail stop. Subsequent ribosomes that also readthrough the standard stop then form a queue behind the first (Figure 2). Once the queue overspills into the main AMD1 open reading frame (ORF) this provides a mechanism to prevent the completion of further protein synthesis.
If this mechanism is correct, it would predict that the longer the tail, a longer queue of ribosomes would be required before they seep into and prevent mainORF expression. Increased stop codon readthrough would then increase queue formation. When the readthrough efficiency was altered, Yordanova et al. found a corresponding drop in reporter expression levels supporting the model.
Nevertheless, the model presented is most likely an oversimplification. Assuming stalled ribosomes release with a rate s, a tail could only form if s < (initiation rate*rate of stop codon readthrough). More research and examples from other genes will be required to further support this model, especially since the formation of ribosome queues have not been seen directly!
As mentioned, AMD1 encodes adenosylmethionine decarboxylase 1, a gene that is important for polyamine synthesis and is highly regulated at the translation stage. Regulating at translation allows more rapid cellular responses as it is the final stage of gene expression. In addition to the tailORF, AMD1 mRNA also has an upstream ORF (uORF) so named because it is upstream of the main ORF (mORF). The uORF encodes the short peptide MAGDIS. Synthesis of MAGDIS stalls the ribosome at the end of the uORF – the longer the ribosome stalls, the less likely the mORF will be translated. The higher the concentration of polyamines, the longer the stall effectively creating a negative feedback loop.
Is AMD1 a special case?
AMD1 is turning out to be remarkably interesting; that with its funky stalling at the end of the uORF and now with its tailORF; but how widespread is this mechanism?
uORFs have been identified in 40-50% of genes, the function for most yet to be determined, but very little is known of tailORFs. The beginning of a new ‘tail’ perhaps.
- Yordanova, M. M. et al. AMD1 mRNA employs ribosome stalling as a mechanism for molecular memory formation. Nat. Publ. Gr. 19, (2018). https://www.nature.com/articles/nature25174